Request: Add a device proactively
Observium team,
I just finished pre-deploying a few routers for a new client, and I'm adding each one to our monitoring platforms in advance. As they're offline, I'll be unable to add them to Observium until they respond to SNMP / ICMP.
Is there a way I could add these proactively, and work around the error that Observium pushes our by default? Ideally, I'd like to add them all now, disable their alerts, and enable them when they're deployed and connectivity is in place.
I see you have a Ignore RRD checkbox on the add device page. I wonder if we could have another that says, "Add even though device is offline"?
As always, thank you for your feedback and time on this.
Regards,
Nate Mellendorf | Network Analyst | Netgain 720 West Saint Germain Street | St. Cloud | MN | 56301 Phone: 320.251.4700 x183 | 877.797.4700 x183 http://www.netgainhosting.com/ www.NetgainHosting.com
_____
The information contained in this email message is for the designated recipient only and may be privileged, confidential, and protected from disclosure. If you have received this message in error, please notify the sender immediately and delete the original. Any dissemination, distribution, copying or other use of this message or any information contained within is strictly prohibited.
Hmm... I'm not sure about this. What do other people think?
adam. On 19/10/2015 16:07:40, Nate Mellendorf nate.mellendorf@netgainhosting.com wrote: Observium team, I just finished pre-deploying a few routers for a new client, and I’m adding each one to our monitoring platforms in advance. As they’re offline, I’ll be unable to add them to Observium until they respond to SNMP / ICMP. Is there a way I could add these proactively, and work around the error that Observium pushes our by default? Ideally, I’d like to add them all now, disable their alerts, and enable them when they’re deployed and connectivity is in place. I see you have a Ignore RRD checkbox on the add device page. I wonder if we could have another that says, “Add even though device is offline”? As always, thank you for your feedback and time on this. Regards, Nate Mellendorf | Network Analyst | Netgain 720 West Saint Germain Street | St. Cloud | MN | 56301 Phone: 320.251.4700 x183 | 877.797.4700 x183 www.NetgainHosting.com [http://www.netgainhosting.com/]
The information contained in this email message is for the designated recipient only and may be privileged, confidential, and protected from disclosure. If you have received this message in error, please notify the sender immediately and delete the original. Any dissemination, distribution, copying or other use of this message or any information contained within is strictly prohibited.
I’m in favor of having an offline device add option. We would be able to add the device to Observium as part of the normal provisioning workflow, instead of post activation.
From: observium [mailto:observium-bounces@observium.org] On Behalf Of Adam Armstrong Sent: Monday, October 19, 2015 8:10 AM To: observium@observium.org Subject: Re: [Observium] Request: Add a device proactively
Hmm... I'm not sure about this. What do other people think?
adam.
On 19/10/2015 16:07:40, Nate Mellendorf <nate.mellendorf@netgainhosting.commailto:nate.mellendorf@netgainhosting.com> wrote: Observium team,
I just finished pre-deploying a few routers for a new client, and I’m adding each one to our monitoring platforms in advance. As they’re offline, I’ll be unable to add them to Observium until they respond to SNMP / ICMP.
Is there a way I could add these proactively, and work around the error that Observium pushes our by default? Ideally, I’d like to add them all now, disable their alerts, and enable them when they’re deployed and connectivity is in place.
I see you have a Ignore RRD checkbox on the add device page. I wonder if we could have another that says, “Add even though device is offline”?
As always, thank you for your feedback and time on this.
Regards,
Nate Mellendorf | Network Analyst | Netgain 720 West Saint Germain Street | St. Cloud | MN | 56301 Phone: 320.251.4700 x183 | 877.797.4700 x183 www.NetgainHosting.comhttp://www.netgainhosting.com/
[cid:image003.png@01D10A55.E8842F90] ________________________________ The information contained in this email message is for the designated recipient only and may be privileged, confidential, and protected from disclosure. If you have received this message in error, please notify the sender immediately and delete the original. Any dissemination, distribution, copying or other use of this message or any information contained within is strictly prohibited.
The way we currently handle this, since all of the devices require dns, is to pre-provision the device on a dhcp network and set dns to the temporary dhcp address. We can do the full config on the on the device, ship it, and when it is in its new location (or while it is being shipped) update dns to the new static IP you pre-provisioned and then when it is plugged in, all should be good with the world. In theory. Most of the time.
Robbie Wright Siuslaw Broadband https://siuslawbroadband.com 541-902-5101
On Mon, Oct 19, 2015 at 8:14 AM, Matt Ladewig matt@univoip.com wrote:
I’m in favor of having an offline device add option. We would be able to add the device to Observium as part of the normal provisioning workflow, instead of post activation.
*From:* observium [mailto:observium-bounces@observium.org] *On Behalf Of *Adam Armstrong *Sent:* Monday, October 19, 2015 8:10 AM *To:* observium@observium.org *Subject:* Re: [Observium] Request: Add a device proactively
Hmm... I'm not sure about this. What do other people think?
adam.
On 19/10/2015 16:07:40, Nate Mellendorf < nate.mellendorf@netgainhosting.com> wrote:
Observium team,
I just finished pre-deploying a few routers for a new client, and I’m adding each one to our monitoring platforms in advance.
As they’re offline, I’ll be unable to add them to Observium until they respond to SNMP / ICMP.
Is there a way I could add these proactively, and work around the error that Observium pushes our by default?
Ideally, I’d like to add them all now, disable their alerts, and enable them when they’re deployed and connectivity is in place.
I see you have a Ignore RRD checkbox on the add device page. I wonder if we could have another that says, “Add even though device is offline”?
As always, thank you for your feedback and time on this.
Regards,
*Nate Mellendorf **|* *Network Analyst **|* *Netgain* 720 West Saint Germain Street | St. Cloud | MN | 56301 Phone: 320.251.4700 x183 *|* 877.797.4700 x183 *www.NetgainHosting.com* http://www.netgainhosting.com/
The information contained in this email message is for the designated recipient only and may be privileged, confidential, and protected from disclosure. If you have received this message in error, please notify the sender immediately and delete the original. Any dissemination, distribution, copying or other use of this message or any information contained within is strictly prohibited.
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
We'll have 0.0 info about the device though, and we won't even know if the SNMP community is correct, ever -- if it's not, device will stay offline forever.
If it's useful to people I don't mind supporting it, although it's obviously have to be a specific override that is not default :)
Tom
On 2015-10-19 17:14, Matt Ladewig wrote:
I’m in favor of having an offline device add option. We would be able to add the device to Observium as part of the normal provisioning workflow, instead of post activation.
*From:*observium [mailto:observium-bounces@observium.org] *On Behalf Of *Adam Armstrong *Sent:* Monday, October 19, 2015 8:10 AM *To:* observium@observium.org *Subject:* Re: [Observium] Request: Add a device proactively
Hmm... I'm not sure about this. What do other people think?
adam.
On 19/10/2015 16:07:40, Nate Mellendorf <nate.mellendorf@netgainhosting.com <mailto:nate.mellendorf@netgainhosting.com>> wrote: Observium team, I just finished pre-deploying a few routers for a new client, and I’m adding each one to our monitoring platforms in advance. As they’re offline, I’ll be unable to add them to Observium until they respond to SNMP / ICMP. Is there a way I could add these proactively, and work around the error that Observium pushes our by default? Ideally, I’d like to add them all now, disable their alerts, and enable them when they’re deployed and connectivity is in place. I see you have a Ignore RRD checkbox on the add device page. I wonder if we could have another that says, “Add even though device is offline”? As always, thank you for your feedback and time on this. Regards, *Nate Mellendorf **|****Network Analyst **|****Netgain** *720 West Saint Germain Street | St. Cloud | MN | 56301 Phone: 320.251.4700 x183 *|* 877.797.4700 x183 *www.NetgainHosting.com* <http://www.netgainhosting.com/> ------------------------------------------------------------------------ The information contained in this email message is for the designated recipient only and may be privileged, confidential, and protected from disclosure. If you have received this message in error, please notify the sender immediately and delete the original. Any dissemination, distribution, copying or other use of this message or any information contained within is strictly prohibited.
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
I agree. solid idea. also on the day that the devices are due to come online, if your on the field starting things up, getting alerts on your phone about them coming up is a great comfort/benefit returning to your desk, it is easy to see which device you haven't connected yet or properly.
In an enterprise, the snmp community string should the be known. but still the ordinary device discovery will occur when pings start coming back.
thanks
Peter Hine Senior Technical Support Engineer (Servers) FCoA ITS peter.hine@familycourt.gov.au
********************************************************************** The information contained in this e-mail (including any attachments) is for the exclusive use of the addressee. If you are not the intended recipient please notify the sender immediately and delete this e-mail. It is noted that legal privilege is not waived because you have read this e-mail. **********************************************************************
Its a quick'n'silly workaround but with a job-scheduler -- couldnt you just schedule the appropriate "add_device.php"-commands? The myriad of flavours of task-scheduling should allow all kinds of flexibility here, and its easy to test, easy to log, easy to follow up...
Observium already uses Cron-entries for polling and housekeeping - there is a way in there already.
The other suggestions of adding a device as "ignored" seems good, but it strikes me as a limited use, single-purpose tool. I'd probably end up adding devices and then forgetting about them..
Meaning visualizing and/or keeping track of "idle/pending/ignored/temporary down/in maintainance" would also have to be adressed. Keeping track of too many shades of colour on the Overview-page isnt good either.
/Paal R/
I'm actually thinking of some kind of "discovery" queue, where a bunch of devices can be added and are then "tested" at the end of each polling cycle to see if they exist. If they do, they're discovered and added, if not, ignored for later.
Putting them directly in to the device table seems messy, especially since we've not verified their authentication information or worked out what OS they are at that point.
adam. On 20/10/2015 14:45:40, Pål Rydningen paal.rydningen@gmail.com wrote: Its a quick'n'silly workaround but with a job-scheduler -- couldnt you just schedule the appropriate "add_device.php"-commands? The myriad of flavours of task-scheduling should allow all kinds of flexibility here, and its easy to test, easy to log, easy to follow up...
Observium already uses Cron-entries for polling and housekeeping - there is a way in there already.
The other suggestions of adding a device as "ignored" seems good, but it strikes me as a limited use, single-purpose tool. I'd probably end up adding devices and then forgetting about them..
Meaning visualizing and/or keeping track of "idle/pending/ignored/temporary down/in maintainance" would also have to be adressed. Keeping track of too many shades of colour on the Overview-page isnt good either.
/Paal R/ _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
Yes!
I actually had a "discovery queue" in mind for longer; we should not try to add new devices on every lldp/cdp/... arp ospf bgp whatever run - just add those things into a queue and try to add them later. That way we also don't try to rediscover a "new" VM every time we get a VM list (vmware/libvirt is in discovery, proxmox is in poller!), rather do this out of band.
We'd keep failed ones in there, this is needed for the feature below too, so if a VM is not able to be added, we should not try again before <configurable timeout>. This matches perfectly with the need below - try re-adding the device periodically that is in the queue.
Now just... to make it so...
On 20/10/2015 17:45, Adam Armstrong wrote:
I'm actually thinking of some kind of "discovery" queue, where a bunch of devices can be added and are then "tested" at the end of each polling cycle to see if they exist. If they do, they're discovered and added, if not, ignored for later.
Putting them directly in to the device table seems messy, especially since we've not verified their authentication information or worked out what OS they are at that point.
adam.
On 20/10/2015 14:45:40, Pål Rydningen paal.rydningen@gmail.com wrote:
Its a quick'n'silly workaround but with a job-scheduler -- couldnt you just schedule the appropriate "add_device.php"-commands? The myriad of flavours of task-scheduling should allow all kinds of flexibility here, and its easy to test, easy to log, easy to follow up...
Observium already uses Cron-entries for polling and housekeeping - there is a way in there already.
The other suggestions of adding a device as "ignored" seems good, but it strikes me as a limited use, single-purpose tool. I'd probably end up adding devices and then forgetting about them..
Meaning visualizing and/or keeping track of "idle/pending/ignored/temporary down/in maintainance" would also have to be adressed. Keeping track of too many shades of colour on the Overview-page isnt good either.
/Paal R/ _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
participants (7)
-
Adam Armstrong -
Matt Ladewig -
Nate Mellendorf -
Peter.Hine＠familycourt.gov.au -
Pål Rydningen -
Robbie Wright -
Tom Laermans